مشاريع طلاب خريف 2023 - F23
تصميم نظام مساعد لمراجعة الأوراق العلمية في مجلة علمية محكمة باستخدام تقنيات الذكاء الصنعي
Designing AI assisted peer-review system in a refereed scientific journal
Peer-Review system is a tool for controlling the quality of scientific publications in scientific journals and conferences, with huge increase of submissions, the pressure on the review process has increased, which depends basically on human efforts in most stages. In this research, we developed a group of AI tools to assist editors in their tasks during review process, this tools include a multi-label classifier which was created based on the deep learning pretrained model for Arabic language, AraBERT, then it was customized by training on data from peer-reviewed journal, then it was used in developing a reviewers recommender system, which is information retrieval model works on classifying and processing Arabic text semantically, which leads to measure the similarity ratio between a paper and collection of publications by reviewers in the same subject to provide top ten suggestions of reviewers with the closest specialization to the paper. The system also provides additional suggestions as new reviewer among previously published articles. Furthermore, a paper quality prediction model was developed, which is the XGBoost regression model trained on statistical features of both accepted and rejected articles, and final decision prediction regression model was developed based on LightGBM model, which was trained on full text of both accepted and rejected articles, The previous two models represent an additional virtual reviewer to assist editor taking the most proper decision for any paper. The established system was tested on data of the peerreviewed journal, and the classification model reached the values 86.5%, 91.8%, 92.4%, and 91.4% for each of the Accuracy, F1-score, Precision, and Recall, respectively. It was also found that 85% of the reviewers suggested by the system are fully suitable for review process, according to the assessment of the journal’s publishing experts, and by applying the recall@10 scale, it showed that 49% of the suggestions submitted by the recommender system matched the actual reviewers. As well as for quality prediction model which achieved accuracy about 84% for R 2 measure, with error rate RMSE about 0.47. Similarly, the final decision prediction model achieved a value of 82% for R2 and an error rate of 0.46.
إعداد: الطالبة المهندسة رزان حلاق
إشراف: الدكتور محمد بسام الكردي
تصميم نظام مساعد لمراجعة الأوراق العلمية في مجلة علمية محكمة باستخدام تقنيات الذكاء الصنعي
نظام دفع إلكتروني باستخدام تقنية البلوك تشاين
Payment System Using Blockchain Technology
This study aims to develop a national cryptocurrency using blockchain technology to offer fast and secure money transfers service in Syrian Arab Republic, addressing weaknesses in traditional centralized bank-reliant electronic payment systems such as low transparency, high costs and vulnerability. a theoretical study of blockchain technology was conducted. System requirements and diagrams have been determined.
The system was built based on these requirements and diagrams, and launched and tested on our personal computer. The results demonstrate the feasibility of building a cryptocurrency system using blockchain from scratch. The proposed system exhibited transparency with an immutable ledger after reaching consensus between network nodes with proof of work consensus mechanism, with possibility to create a blockchain wallet within minutes and a transaction processing speed of 30 seconds, transactions are made directly from one wallet to another without the need for brokers, along with a high level of security thanks to using SHA 256 hash function, elliptic curve cryptography and elliptic curve digital signature algorithm , the system showed fault tolerance thanks to using pure unstructured peer-to-peer network with topic based publish subscribe model where new nodes have joined the system and synced its ledgers successfully with other nodes and failure of a node didn't affect other nodes. Furthermore, the tested system has low operating costs. However, it is important to acknowledge the limitation of testing the distributed system on a single computer. Real-world scenarios involving a larger network of nodes and varying conditions may present additional challenges.
إعداد: الطالب كنان علي سليمان
إشراف: الدكتورة سيرا أستور
تطبيق الهاتف المحمول القائم على نظام أندرويد للتنبؤ بالأمراض
Android-based mobile application for disease prediction
During the previous period, especially after the Corona pandemic and the curfews that accompanied it and the wheel of life almost stopping, interest increased in employing artificial intelligence in applications for disease prediction, prevention, and treatment methods, and this is what was studied in this thesis, where we developed an application based on artificial intelligence algorithms to predict diseases.
As a result of the appearance of certain symptoms, this application helps doctors or patients alike, especially in emergency cases or residents of villages and remote areas. The focus was on how to show the results of the algorithms in a simple, easy-to-use application that is available to everyone. A database of 4,920 patients from Kaggle was used. After describing it, cleaning it, and preparing it to start training, it was trained on 3 algorithms: (Random Forest, SVM, Decision Tree), and then the results were tested and measured on a section of the data and on external data, and a correct prediction rate of 95.13% was reached using SVM, and from Then he was deported.
The results of the algorithm are transmitted to Flask's web application and by sending HTTP post requests Flask then extracts the disease for symptoms show it on an android-app. This message contributes to providing an effective and reliable solution to assist doctors in early diagnosis and improve the quality of health care. These technologies can also be relied upon to develop effective health policies and disease prevention strategies
إعداد: الطالب حسين غسان الشلوف
إشراف: الدكتور وسيم جنيدي
تطبيق الهاتف المحمول القائم على نظام أندرويد للتنبؤ بالأمراض
تطوير نظام لكشف المواقع المزيفة بالاعتماد على تقنيات الذكاء الصنعي
Developing a fake website detecting system using artificial intelligence techniques
Protecting users from phishing websites has become critically important, as many users download and install software or conduct financial transactions without verifying the authenticity of the website. Users can avoid fraudulent websites by checking for security certificates and trusted signs on the site. The most effective solution lies in relying on anti-malware software and regularly updating programs, as well as avoiding phishing emails and sticking to well-known, reputable websites.
In today's digital world, protecting individuals and companies against phishing websites is essential. Caution must be exercised when opening email links and checking incoming emails. It is also recommended to avoid websites that appear to be official but have an unfamiliar domain extension. Additionally, antivirus software and checking for security certificates on sensitive websites can provide an extra layer of security. Protection against phishing websites is a fundamental part of digital safety and preserving online privacy.
In this research, we selected a dataset [4] that classifies websites based on several parameters indicating whether they are fraudulent or not. We applied the XGBOOST algorithm, which achieved an accuracy rate of 86% in predicting website legitimacy
إعداد: الطالب المهندس لؤي معتز الشيخه
إشراف: الدكتورة سيرا أستور
تطوير نظام لكشف المواقع المزيفة بالاعتماد على تقنيات الذكاء الصنعي
تصميم تطبيق ويب يستخدم تقنيات الذكاء الصنعي للقيام بالحسابات الهندسية
Designing a web application that uses artificial intelligence techniques to do engineering calculations
The human visual perceptual system achieves remarkable performance for objects recognition even in noisy environment or background contrast and the concept of amazing ability for a comprehensive review of the evolution of image processing by have developed especially using convolution neural networks which has led to a revolution in image processing and achieving performance at the human level. The performance of CNN networks has demonstrated the possibility of achieving extraordinary performance in rapid visual recognition. Since the field of civil engineering is far from the field of technology, we will, with these thesis study of History of Engineering and its relationship to programming and using computer vision to identify structural elements, type of supports and elements and using these features to determine the repoct of stable structural element and compare it with previous application. The Yolov8 technology was used to train the neural network on the types of supports and use the elements in each establishment and use the python language to write the programming code for the engineering calculations.
إعداد: الطالبة المهندسة ربى محمد غزول
إشراف: الدكتور محمد مازن المصطفى
تصميم تطبيق ويب يستخدم تقنيات الذكاء الصنعي للقيام بالحسابات الهندسية
تصميم وتنفيذ نظام توصية على موقع ويب للأفلام
Design and Implementation a Recommendation System on Movies Website
Movie viewers often encounter challenges in finding content that aligns with their expectations on movies websites, which negatively impacts the overall user experience.
This project aims to enhance that experience by developing an efficient recommendation system. The system incorporates two models: one based on content-based filtering, and another that combines both collaborative filtering and content-based filtering.
The primary objective of this project is to build a recommendation system capable of delivering accurate, relevant, and personalized suggestions tailored to the user's preferences. Additionally, we will create a website to showcase movie and actor information, along with personalized recommendations through intuitive and user-friendly interfaces, thereby improving the user experience and meeting their needs. The research will begin with an introduction that defines the problem addressed by the study and the objectives we aim to achieve by the end of the research. It will also provide a historical overview of recommendation systems, their origins, and examples from various application fields, particularly in the context of the global movie website industry.
In the first chapter, we will explore movie websites and their requirements, identifying the key factors that differentiate one movie website from another. We will then provide an in-depth explanation of recommendation systems, their types, including the strengths and weaknesses of each system, followed by a comparative analysis. The chapter will conclude with a discussion of the general limitations of recommendation systems.
In the second chapter, we will review a selection of reference studies that have contributed to our understanding of recommendation systems, their types, and algorithms. Additionally, we will highlight related works that served as inspiration for our research.
The third chapter will detail the methods and steps followed during the development stages of the recommendation system, including the techniques and libraries used in the practical implementation, as well as the testing of the results. Finally, in the fourth chapter, we will present the web interfaces that users will interact with.
The study will conclude by discussing future directions for this work and listing the references that informed our research
إعداد: الطالبة لينا محمد عماد موصللي
إشراف: الدكتورة سيرا أستور
تصميم وتنفيذ نظام توصية على موقع ويب للأفلام
بنــاء تطبيـــق ويـــب مع تطبيـــق نقـــال لأتمتـــة معامـــلات العامليــن في الجمعيات والمنظمات المحلية غير الحكومية واستخدام تقنيــات الذكاء الصنعي لتوقع ضغط العمل وحالة العاملين.
Build a web application with a mobile application to automate the transactions of workers in associations and local NGOs and use artificial intelligence techniques to predict work pressure and the status of workers.
This project aims to develop a comprehensive system that includes a web application and a dedicated mobile application for automating the transactions of employees in local non-governmental organizations and associations. This system is centered around improving the efficiency and effectiveness of daily employee operations, saving time and effort and reducing potential human errors. It also provides cost and environmental benefits through reduced printing and paper consumption.
One of the key features of this system is the integration of artificial intelligence (AI) technologies. AI tools will be embedded to provide analytics and forecasts on workload and employee status. For example, the system can analyze attendance and activity data to predict peak periods where employees may require additional support. It can also track performance indicators and generate reports to assist management in making informed decisions.
Additionally, the system offers robust security features to ensure the protection of sensitive employee and transaction data. All data transmitted and received between the applications and servers is encrypted, and two-factor authentication options are provided to ensure only authorized personnel can access the system.
This project aims to enhance collaboration and transparency within associations and non-governmental organizations, contributing to improved work environments and increased job satisfaction among employees. By automating processes and providing advanced analytical tools, these organizations can focus more on achieving their humanitarian and social objectives, rather than being preoccupied with routine administrative operations.
إعداد : الطالب وجدي ميلاد السكاف
إشراف: الدكتورة سيرا أستور
تصميم وتنفيذ تطبيق وب ذكي لعلاقة المدرسة مع الطالب
Designing and implementation a smart web application for the school-student relationship
Education is at the present time one of the most important requirements, so schools seek to provide the best services to improve the level of their students and provide assistance to them whether they are inside or outside it.
Many students in the middle school also suffer from difficulty in determining their field of study in the high school, because of their young age and lack of proper guidance from their parents. Therefore, in this research, students will be classified using machine learning algorithms, while proposing a recommendation system to improve the level of students at this stage.
This study is divided into eight chapters and begins with the first chapter (theoretical study): which provides a comprehensive theoretical study of this field with a detailed explanation of machine learning algorithms. While in the second chapter (the reference study), we present a summary of five previous studies closely related to the subject of our study, highlighting the most important negative and positive aspects of each of them. In the third chapter (the proposed system), we explained the structure of the system using box diagrams. In Chapter Four (Analytical and Design Study), we analyzed the system in detail, starting with defining its functions and moving on to studying the functional and non-functional requirements. We also mentioned the tools and technologies used in its development, in addition to building use case diagrams and analyzing ERD diagrams. Then in the fifth chapter (Achievement): We explained in detail the training of the model used to classify students and the adoption of the random forest algorithm in classification. In Chapter Six (Results and Tests): The system was fully built and all scenarios that the user could perform were explained. Then we evaluated the system in Chapter Seven (Evaluation and Comparison): by mentioning the advantages of the system and comparing its results from previous studies. As we mentioned in Chapter Eight (Conclusion and Future Prospects): The points we aspire to improve the performance of the completed system.
إعداد: الطالبة رزان محمد اختريني
إشراف: الدكتور محمد بسام الكردي
تصميم وتنفيذ تطبيق وب ذكي لعلاقة المدرسة مع الطالب